RE: -fno-inline -O1 breaks ntohs()

2006-09-17 Thread Danny Smith
> > > >! #if defined __OPTIMIZE__ && !defined __NO_NOINLINE__ #if defined __OPTIMIZE__ && !defined __NO_INLINE__ Rule No 1: The obvious is not always not what it appears to be. Danny -- Unsubscribe info: http://cygwin.co

Re: -fno-inline -O1 breaks ntohs()

2006-09-17 Thread Christopher Faylor
On Mon, Sep 18, 2006 at 10:04:45AM +1200, Danny Smith wrote: > >> Testcase: >> -- >> #include >> >> int main() >> { >> volatile short s = 33, t; >> >> t = ntohs(s); >> return 0; >> } >> -- >> >> > i686-pc-cygwin-gcc -fno-inline -o t

RE: -fno-inline -O1 breaks ntohs()

2006-09-17 Thread Danny Smith
> Testcase: > -- > #include > > int main() > { > volatile short s = 33, t; > > t = ntohs(s); > return 0; > } > -- > > > i686-pc-cygwin-gcc -fno-inline -o test test.c > i686-pc-cygwin-gcc -O1 > > -fno-inline -o test test.c > /tm

-fno-inline -O1 breaks ntohs()

2006-09-17 Thread Stefan Traby
Hi! Testcase: -- #include int main() { volatile short s = 33, t; t = ntohs(s); return 0; } -- > i686-pc-cygwin-gcc -fno-inline -o test test.c > i686-pc-cygwin-gcc -O1 -fno-inline -o test test.c /tmp/ccswmuY9.o:test.c:(.text+0x27):